Abstract
Recent tsunami disasters, such as the 2010 Chilean and 2011 Tohoku events, demonstrate the significant life loss that can occur from tsunamis. Many coastal communities in the world are threatened by near-field tsunami hazards that may inundate low-lying areas only minutes after a tsunami begins. Geospatial integration of demographic data and hazard zones has identified potential impacts on populations in communities susceptible to near-field tsunami threats. Pedestrian-evacuation models build on these geospatial analyses to determine if individuals in tsunami-prone areas will have sufficient time to reach high ground before tsunami-wave arrival. Areas where successful evacuations are unlikely may warrant vertical-evacuation (VE) strategies, such as berms or structures designed to aid evacuation. The decision of whether and where VE strategies are warranted is complex. Such decisions require an interdisciplinary understanding of tsunami hazards, land cover conditions, demography, community vulnerability, pedestrian-evacuation models, land-use and emergency-management policy, and decision science. Engagement with the at-risk population and local emergency managers in VE planning discussions is critical because resulting strategies include permanent structures within a community and their local ownership helps ensure long-term success. We present a summary of an interdisciplinary approach to assess VE options in communities along the southwest Washington coast (U.S.A.) that are threatened by near-field tsunami hazards generated by Cascadia subduction zone earthquakes. Pedestrian-evacuation models based on an anisotropic approach that uses path-distance algorithms were merged with population data to forecast the distribution of at-risk individuals within several communities as a function of travel time to safe locations. A series of community-based workshops helped identify potential VE options in these communities, collectively known as "Project Safe Haven" at the State of Washington Emergency Management Division. Models of the influence of stakeholder-driven VE options identified changes in the type and distribution of at-risk individuals. Insights from VE use and performance as an aid to evacuations from the 2011 Tohoku tsunami helped to inform the meetings and the analysis. We developed geospatial tools to automate parts of the pedestrian-evacuation models to support the iterative process of developing VE options and forecasting changes in population exposure. Our summary presents the interdisciplinary effort to forecast population impacts from near-field tsunami threats and to develop effective VE strategies to minimize fatalities in future events.